Stephen Jerge Ψ

Clinical-Science–Track Researcher | Psychological Research Developer

I build small, practical tools that support clinical and mental-health research—one-command runs that clean/de-ID data, flag issues, and leave a manifest.

See it fast: (Demo link here when recorded) — PRDT run with sample PHQ-9/GAD-7 data.

Try it:

python3 -m venv .venv && source .venv/bin/activate
pip install https://github.com/stephenmjerge/psych-research-data-toolkit/releases/latest/download/prdt-0.1.4-py3-none-any.whl
prdt --config https://raw.githubusercontent.com/stephenmjerge/psych-research-data-toolkit/main/configs/anxiety.toml --outdir /tmp/prdt-demo

Outputs: clean/anonymized CSV, alerts.json (missingness/PHI/reliability), plots, and run_manifest.json.

---

Core Projects (clinician-facing first)

1. PRDT — Psych Research Data Toolkit
   Clean + de-ID REDCap/Qualtrics exports; emit PHI quarantine, missingness/reliability alerts, plots, and a reproducible manifest.

2. MPAL — Mini Psychological Assessment Lab
   Scores PHQ-9, GAD-7, PCL-5, BDI-II; outputs severity labels, simple trends, and structured summaries for research/educational use.

3. PRWO — Psych Research Workflow Organizer
   Weekly checklist + scripts that keep roadmaps/PortfolioHub/readmes in sync; posters workspace for artifacts before OSF upload.

4. ADOIF — Article / DOI Fetcher
   Pulls PDFs/metadata and organizes a review library for lit sweeps and faculty targeting.

5. RL-Sim — Reinforcement Learning Simulation
   Tiny reward-learning simulations to link symptom severity and decision-making.

---

Why this matters (for clinics/research)

• Safer handling: HMAC anonymization + PHI guardrails; data dictionaries and alerts to catch issues early.
• Reproducible runs: manifests and configs per execution; tests run via python -m pytest tests (PRDT).
• Clinician-ready outputs: cleaned CSVs, severity summaries, and plots that slot in after REDCap/Qualtrics and before SPSS/R.